The present invention relates generally to audio signal processing circuits, and more specifically to audio signal processing circuits for reducing noise in audio signals.
Telephone communications are made easier and sometimes safer through the use of so-called xe2x80x9chands-freexe2x80x9d microphones. Such microphones are typically located remote from the talker or talkers and thus pick up many times more noise and echo than with typical a hand-held unit. This problem is typically compounded in cellular telephone communications systems located in a moving vehicle due to added sources of noise such as road noise, for example.
Noise reduction units are known and are operable to enhance the sound quality and intelligibility of voice recognition systems by providing up to 15-20 dB of effective noise reduction. However, known noise reduction systems have drawbacks associated therewith. For example, some such noise reduction systems using digital signal processing technology create an artificial quality to the reproduced voice and are costly to implement. Known dynamic noise reduction (DNR) circuits, on the other hand, do not appear to work as well in very noisy environments, as xe2x80x9cfuzzballsxe2x80x9d of modulated noise tend to appear in the received signal.
What is therefore needed is an improved noise reduction system operable to enhance sound quality and intelligibility in telephone communication systems employing hands-free microphones, wherein such a noise reduction system is inexpensive and easy to implement, and provides a significant improvement in the signal to noise ratio while minimizing audible distortion.
The present invention addresses the foregoing drawbacks of prior art distortion limiting amplifiers. In accordance with one aspect of the present invention, a signal processing circuit for reducing noise in an audio signal comprises a pre-emphasis circuit having an input receiving an audio signal and an output providing a pre-emphasized representation of the audio signal, the audio signal including a higher amplitude audio component and a lower amplitude noise component, an amplifier including a first input coupled to the pre-emphasis circuit output and an output, the amplifier defining a smooth transfer function providing higher gain to the higher amplitude audio component and lower gain to the noise component of the audio signal, and a de-emphasis circuit having an input connected to the amplifier output and an output providing a de-emphasized representation of the audio signal processed by the amplifier in accordance with the transfer function.
In accordance with another aspect of the present invention, an amplifier circuit for reducing noise in an audio signal comprises a non-inverting input receiving an audio signal, the audio signal including a higher amplitude audio component and a lower amplitude noise component, an inverting input, an output, and feedback circuitry connected between the output and the inverting input. The feedback circuit defines an amplifier transfer function having a smooth non-linear characteristic providing higher gain to the higher amplitude audio component and lower gain to the lower amplitude noise component.
In accordance with a further aspect of the present invention, an amplifier circuit for reducing noise in an audio signal comprises a non-inverting input connected to ground potential, an inverting input, an output, and feedback circuitry connected between the output and the inverting input, wherein the feedback circuitry receives an audio signal including a higher amplitude audio component and a lower amplitude noise component. The feedback circuitry defines an amplifier transfer function having a smooth non-linear characteristic providing higher gain to the higher amplitude audio component and lower gain to the lower amplitude noise component.
One object of the present invention is to provide an audio signal processing circuit providing a non-linear gain proportional to signal amplitude.
Another object of the present invention is to provide such a circuit wherein the non-linear gain is smooth and bi-directional to thereby minimize audible distortion of the processed audio signal.
Yet another object of the present invention is to provide such a circuit further having the audio signal pre-processed by a pre-emphasis circuit and post-processed by a de-emphasis circuit to thereby further minimize audible distortion of the processed audio signal.
These and other objects of the present invention will become more apparent from the following description of the preferred embodiment.